Roles and Responsibilities

• Design and develop the architecture for vSLAM systems to support multi-camera, high-resolution setups.
• Optimize the visual feature extraction process across multiple cameras.
• Leverage CUDA and Tensor Core optimization techniques to accelerate vSLAM processing.
• Develop advanced algorithms to analyze and validate the integrity of data from various sensors (e.g., IMU, Encoder).
• Study and optimize existing controller hardware to improve overall system performance, ensuring stability and efficiency.
• Continuously improve the efficiency, speed, and scalability of vSLAM algorithms, adapting them to meet real-time system requirements.
• Develop logical test plans and data collection strategies to evaluate the effectiveness and performance of vSLAM solutions against defined goals and key metrics, ensuring system robustness.

Requirements:

• Master's or Ph.D. degree in Electrical, Mechanical, or Computer Engineering or a relevant discipline, with more than five years of industry experience.
• Proficiency in C++, with substantial experience in scripting such as Python.
• Extensive expertise in 3D computer vision, multi-view geometry, and SfM/SLAM.
• Experience with CUDA and Tensor Cores for performance optimization on GPUs.
• Knowledge of sensor calibration, data synchronization, and handling data from multi-sensor systems.
• Mathematical foundation, particularly in multi-view geometry, linear algebra, and optimization.
• Familiarity with open-source optimization libraries such as g2o, GTSAM, and Ceres.
• Expertise in optimization, numerical linear algebra, probabilistic estimation, and sensor fusion.
• Experience with debugging, profiling, and optimizing code for performance on both CPUs and GPUs.